Scientific Articles
Below you will find links to a variety of scientific articles published in peer-reviewed journals. These articles will hopefully serve as a starting point for understanding the current state of AI literature as it applies to healthcare and, more specifically, to Emergency Medicine.
Please use the drop-down menus below to view studies.
Background studies
The studies below offer a broad overview of artificial intelligence in healthcare.
A short guide for medical professionals in the era of artificial intelligence
Published in Nature. Defines AI and its different levels, discusses what machine learning is as well as its subtypes, and also compares machine and deep learning. A great introduction towards better understanding what AI is and can be.
Artificial intelligence in healthcare: Past, present and future
Published in BMJ's Stroke and Vascular Neurology. Reviews the motivations of applying AI in healthcare, data types that have be analyzed by AI systems, the mechanisms that enable AI systems to generate clinical meaningful results, and the disease types that the AI communities are currently tackling.
The practical implementation of artificial intelligence technologies in medicine
Published in Nature Medicine. Review discussing the practical issues surrounding the implementation of AI in medicine as well as the current regulatory environment within the United States and globally.
Artificial intelligence in healthcare
Published in Nature Biomedical Engineering. Discusses different AI technologies and summaries the economic, social, and legal implications of AI in healthcare.
Artificial intelligence in medicine
Published in Metabolism. Historical overview of technology in healthcare culminating in a discussion surrounding AI, what it is, its current utility, and how it may be used in the future.
AI in EM (Broad)
The below studies offer a look into the potential applications of AI in Emergency Medicine highlighting current and future use cases, ares for improvement, and both current and anticipated challenges.
Artificial intelligence in emergency medicine: A scoping review
Published in JACEP Open. A scoping review of 150 articles on AI that are examined and characterized with a focus placed on EM. Offers a detailed overview of the current literature available.
Artificial Intelligence in Emergency Medicine: Benefits, Risks, and Recommendations
Published in the Journal of Emergency Medicine. Extensive discussion on the various aspects of AI including legal issues, privacy and confidentiality, big data, incorporating AI into medical education, etc. Also provides a list of recommendations based on their analysis of the role of AI in EM.
Artificial Intelligence in Emergency Medicine: Surmountable Barriers With Revolutionary Potential
Published in the Annals of Emergency Medicine. More nuanced discussion surrounding the adoption of AI in EM, the current regulatory and legislative barriers, as well as the challenge of integrating AI into current workflows.
Published in BMC Health Services Research. Literature review on AI and work design in the ED. Current literature shows that AI support is mostly utilized in the triage of patients but there is ample evidence that AI-tools could also improve the clinical decision-making process.
Applications of AI
The following studies discuss different applications of AI in healthcare and represent some of the first studies to serve as proof=of-concepts and validation studies. They are further differentiated into those directly applying to EM and those studies performed within other adjacent fields.
Applications of AI in EM
Published in JACEP Open. Deployed a previously validated AI algorithm to assist in the interpretation of chest radiographs during the first wave of the pandemic. First published literature on an AI tool that impacted clinical decision making in the ED setting.
EM-Adjacent Applications of AI
Machine learning for ECG diagnosis and risk stratification of occlusion myocardial infarction
Published in Nature Medicine. First observational cohort study to develop machine learning models for the ECG diagnosis of OMI. Derived and externally validated an intelligent model that outperformed practicing clinicians.
Published in Circulation: Arrhythmia and Electrophysiology. Evaluated a cardiologist-designed AI algorithm the examine ECGs for evidence of LVSD in dyspneic ED patients. Found to out perform NT-proBNP.
AI in Healthcare Admin
The AI-Enhanced Future of Health Care Administrative Task Management
Published in NEJM Catalyst. Commentary piece discussing the creation of a five step framework (ISUMO) to better understand and implement AI-enhanced task management in health care.